question

petajoule avatar image
petajoule asked

2-signal (4-wire) BMS on Quattro woes

Hi,

I have a 16-cell 90Ah LiFePO4 battery using LiPro-1 (https://ecs-online.org/BMS/LiPro1-1:::2_7.html ) BMS. This BMS has a LVP (low voltage protection) and a OVP (overvoltage protection) circuit, so I thought I'll let it talk to my Quattro.

Connected LVP circuit to AUX1, OVP circuit to AUX2.

Installed the 2-signal Assistant with the following parameters:

"Battery empty when AUX1 open"
"battery full when AUX2 open"
"When AUX2 open, disable charger and do not adapt SOC"


So... after starting that assistant (and blackout :-)) I'm running now on battery only. Somehow the quattro does not want to take any juice from mains anymore.

At the current rate, I will have the next blackout in about 1 hour, as wife is baking...

Does anyone have an idea what could be going on here? Mains is on ofc, checked, both circuits (LVP/AUX1 and OVP/AUX2) are closed.


BMSAssistants
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

3 Answers
kai avatar image
kai answered ·

Are you able to verify the intended behaviour of the Quattro if you were to remove the BMS from the loop and false feed the AUX1 and AUX2 manually (open = nothing at relevant terminals, closed = relevant terminals jumpered with wire).

Also, the interface is supposed to be implemented using either voltage free contacts or an optocoupler. If the BMS is driving the interface with a logic level output you may have to put an interface circuit in to make the controls amenable to the Quattro.

1 comment
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

petajoule avatar image petajoule commented ·

Testing/seeing what the Quattro thinks is the state of the AUX inputs would be a great thing to do. Unfortunately every time I use VEConfigure (3 by now), I feel like the 90ies want their software back.

I'm missing there somewhere AUX1 open/closed, AUX2 open/closed indicator, and actually also the indicator for the other gizmos in the Quattro, like K1 on/off etc.
I mean the information must be available to the Quattro.

All I see, is AUX 0, next to AC ignore 0, which is not only non-informative - it also does not provide the "What's this?" documentation.


I figured the problem might be because the BMS is keeping a closed loop of the battery voltage (probably designed to drive relais) and the Quattro doesn't like that. A pity, I had hoped to spare the two relais - or optocoupler.


0 Likes 0 ·
petajoule avatar image
petajoule answered ·

addendum:


It all got worse - much worse. Fortunately I was monitoring the system closely, so I could prevent serious damage to the cells.


I waited observing the system discharging. After - according to the Quattro - the SOC of the battery fell to 15%, several cells started failing rapidly, LVP circuits opened in 3-4 cells, quattro (and it's 2-signal assistant) were unimpressed.

I therefore changed the assistant parameter from "turn charger off" to "force float", but had no high hopes as that is the behavior for "battery full" condition. After the usual blackout when you update the assistant, the Quattro tried to bring up the power. Twice. And it failed twice, probably ignoring the AC main.

*BIG SIGH*

So I also rapidly removed the assistant from the quattro, blackout again, but then fortunately everything started to work again. AC main was not ignored anymore battery started charging, so cell damage could be averted. I hope.

Overall not a very pleasant experience.

2 comments
2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Matthias Lange - DE avatar image Matthias Lange - DE ♦ commented ·

1. Programing and testing should be done without the the house connected. If the system works as it should you can connect a single load for testing and if this works you can connect the house.

2. Maybe there was a problem during the first programing causing the misbehavior.

0 Likes 0 ·
Show more comments
kai avatar image
kai answered ·

hi again

I can only re-suggest the option of testing at the interface. Set up the battery conditions such that you can either:

1) know the battery is full and therefore you can test inverter on/off function

2) know the battery needs charging and therefore you can test charger on/off function

Use wire jumper accordingly. Of course you'll have to monitor the BMS separately at this point and be prepared to manually intervene if the voltage goes out of spec.

You can also test the BMS output - if you can see a voltage across the LVP or OVP outputs then it is not a voltage free output (make sure you check both states). If its a relay or opto output you shouldn't see a voltage (but can see a change in resistance when the BMS toggles the state).


Need to do the testing at a lower level before moving onto whole of system tests.


Cheers


2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.